RU2691659C1 - Method for non-invasive identification of objects by nmr spectra and device for its implementation - Google Patents

Abstract

FIELD: physics.SUBSTANCE: use: for examination and non-invasive identification of visually indistinguishable samples by nuclear magnetic resonance. Summary of invention consists in the fact that device for NMR detection of samples differing by their characteristics from reference, includes two inductance identical inductance coils connected in series to a signal receiver and configured to receive a reference sample in one of them, into a second sample, wherein one of the coils is connected by one end to one of the input ends of the signal receiver, the other – to one of the ends of the second coil, the remaining end of the second coil is connected to the other input end of the signal receiver, the coils are configured to switch modes of operation, at which induction EMF induced at the coil ends are summed or subtracted, and the device is configured to be placed in a homogeneous magnetic field with perpendicular location of the coil axes relative to the direction of the given field. Method of detecting samples, differing by their characteristics from reference, includes arrangement of reference and analysed samples in the above device with subsequent arrangement of coils with samples in homogeneous magnetic field with provision of perpendicular location of coils axes to given field, followed by radio-frequency excitation of resonant spins by NMR simultaneously in both samples with subsequent recording of NMR signals simultaneously from both samples to said excitation, and the sample, differing by its physical and chemical characteristics from the reference, is determined based on the results of comparing the NMR spectra of the obtained signals.EFFECT: high accuracy of identifying the analysed object.8 cl, 10 dwg

Description

Technical field

The invention relates to the study and non-invasive identification of visually indistinguishable samples in order to identify samples that differ in their physico-chemical characteristics from the reference. Comparison of the investigated and reference samples is made according to their NMR spectra and the results of processing the obtained data. Based on them, differences in spectral characteristics are revealed - in peak intensities and line widths, the presence of additional signals due to impurities, etc.

The invention can be widely used in the identification of counterfeit objects, including produced industrially, for example, for liquids poured in the same containers. In this case, the inventive method allows to detect differences in the volume of the liquid, the content of impurities in it, etc. At the same time, in order to conduct studies of compared objects, they must be available for registration by NMR - in particular, they must be of a size that is acceptable for their placement in magnetic resonance equipment, and must not contain ferromagnetic inclusions.

The level of technology

The claimed solution involves the use of a non-invasive method, namely, the method of nuclear magnetic resonance (NMR), with which you can record signals from an isotope with a magnetic moment, which is possible if its nuclear spin I is different from zero. Of greatest interest for the proposed method are substances containing isotopes of atoms that make up organic compounds. In addition to protons ¹ H (I = 1/2), it can be its heavy isotope - deuterium ² H (1 = 1), as well as carbon isotopes - ¹³ C (I = 1/2) and oxygen ¹⁷ O (I = 5 / 2).

If during the registration of proton spectra it is not possible to detect differences in the molecular structure and impurity content, then the registration of these isotopes may reveal differences in the isotopic composition of the samples. This is true, in particular, for water-containing samples, since the content of deuterium depends on the geographical latitude of the water source, or for alcoholic beverages, since the content of ¹³ C isotope in alcohol depends on the raw materials from which this product is made (grain, grapes, cane, etc. ) (Martin ML, Martin GJ NMR Basic Principles and Progress. V. 23, Springer-Verlag, Berlin-Heidelberg. 1990). Due to the relevance of such measurements, a regulated procedure has been developed based on mass spectrometry (GOST 32710-2014. Alcoholic products and raw materials for their production. Identification. Method for determining the 13C / 12C isotope ratio of spirits and sugars in wines and wort. Moscow. Standardinform. 2014) . This method is not readily available. But the main problem is its invasiveness, since only a fragment of the sample is analyzed, because of which the packing of the sample is broken.

The prior art uses the use of NMR and MRI for the study of objects (Non-Medical Applications of NMR and MRI, http://www.magnetic-resonance.org/ch/19-01.html), see also (Schmidt SJ, Sun X., Litchfield JB, Eads TM Applications of magnetic resonance imaging in food science // Critical Reviews in Food Science and Nutrition (1996. V. 36 (4) P / 357-385), but there are no sources of information in which describes the use of NMR and MRI for simultaneous non-invasive analysis of the test and reference samples in order to identify differences in both the volume of the liquid and the content of impurities in it.

In particular, high-resolution NMR spectroscopy methods are known from the state of the art (eg, Marcone MF et al., NMR) technology // Food Research International, 2013, 51 (2): 729-747 ), which can be implemented on magnetic resonance equipment - high-resolution NMR spectrometer, provided that at least one sample under study can be placed in its working area, or magnetic resonance imaging, where it is possible to place objects commensurate with the size of a person. Then it is possible to successively measure the NMR signals from each sample separately, followed by the calculation of the spectra and the calculation of their ratio. It is implemented as follows. Object A is placed in the receiving coil of magnetic resonance equipment and the NMR spectrum S _{A is} recorded. Then, instead of sample A, sample B is placed in the same receiving coil, and with the same NMR signal acquisition parameters, an NMR spectrum is obtained from sample B – S _B. According to the data obtained, the ratio S _A / S _B or (S _A -S _B ) / (S _A + S _B ) is calculated.

The disadvantage of the method is that the registration of the NMR signals for samples A and B are separated in time, so it is not always possible to provide the same shooting conditions for each of the samples. This is especially inconvenient when registering weak signals, where multiple signal accumulation may be required to obtain an acceptable signal-to-noise ratio and the spectrum acquisition time may be significant. In this case, there may be problems with temperature stabilization of circuit parameters, stability of generator power, etc. Changing the temperature of the sample and circuit, generator power and other parameters from time to time create non-identical conditions for the registration of spectra for objects recorded at different time intervals. Therefore, the results of measurements of S _A / S _B are unreliable.

The method of local spectroscopy is also known (Kimmich R., Hoepfel D. Volume-selective multipulse spin-echo spectroscopy // J. Magn. Reson. 1987. V. 72 (2). P. 379-384). In this case, both samples are placed in a common receiving coil, and they are there for the duration of the measurements. The idea of local spectroscopy is to accept an NMR signal from a given coil, i.e. from both samples, but before that conduct RF excitation of the spins only in the local region. At first such area is the location of sample A, from which the spectrum S _A is recorded. Then local excitation of object B is performed, from which the spectrum of S _B is recorded.

Spatial-selective pulses are used to implement the method of local spectroscopy — RF excitation of the spin system is performed in the presence of inhomogeneous (gradient) magnetic fields. Due to the special assignment of the form of the RF pulse, its duration, filling frequency, as well as the amplitude of the gradient field, the RF spins are excited only in the region of a given size and localization. The pulse shape and its duration τ determine the effective frequency spectrum of the pulse - Δf = k / τ, where k is a coefficient (from 1 to 10), depending on the shape of the pulse. For the local excitation of spins with a gyromagnetic ratio γ in the Δz layer at a distance Z from the center of the magnet, a gradient G _z and an RF pulse with a spectrum Δf and a filling frequency f are applied simultaneously. This frequency is set according to the relation γG _z Z = ff ₀ , where f ₀ is the Larmor frequency measured at the center of the magnet, and G _{z is} set from the relation γG _z Δz = Δf. By applying successively pulses with different filling frequencies and gradient values along the remaining orthogonal directions, an excited volume of Δx × Δу × Δz in the required localization (X, Y, Z) is selected. For this method, the problem of the separate recording of signals from different objects is not so relevant, since the accumulation of NMR signals from different objects can be carried out by their alternate excitation.

The disadvantage of this method compared to conventional spectroscopic methods is the need for additional instrumentation. This applies to gradient coils and current sources for them, as well as equipment for the formation of pulses of complex shape in order to obtain a narrow spectrum of Δf. In addition, the application of the technique of multiple spin echo increases the measurement time - because a three-pulse sequence is used for registration, where delays are inserted between the pulses. In conventional spectroscopic methods for the excitation of spins, there is no need to use the technique of multiple spin echo, and therefore a single pulse can be used without special delays.

In addition, the method places increased demands on the power of the RF system — its ability to realize short and powerful RF pulses. Their duration should be less than the relaxation time of the spin system, and the power should be sufficient to provide a large deflection angle of the magnetization vector FA (Flip Angle) - 90 and 180 degrees, without which the implementation of the multipulse spin echo technique, which is used for local spectroscopy, is impossible. The greatest difficulties in the implementation of the spin echo method arise with short relaxation times of the spin system and large sample sizes. The shorter the relaxation time of the spins, the shorter the duration of the RF pulse τ for its excitation should be. And since FA = γB ₁ τ, where B ₁ is the amplitude of the RF field, which excites the spins, for a given value of FA, the power of the transmitter that generates RF pulses with amplitude B ₁ is inversely proportional to this duration.

The larger the sample size, the larger the transmitting coil size must be in order to create a uniform RF field in the entire sample zone. And with increasing coil size, more current is required through it in order to create the required RF field B ₁ at the center of the coil. As a result, requirements for RF pulses (duration or power) are not always realizable, especially on typical MRI equipment.

In conventional spectroscopic methods, pulses with a small deflection angle FA can be used, so the RF transmitter power requirements are very mild.

In addition, the use of gradient pulses can cause a destructive effect on the studied samples. This is due to the fact that their use is accompanied by acoustic noise due to the mechanical vibration of the current coils with which these fields are created. Vibration is caused by the mechanical displacement of the conductor through which current flows, changing in time, under the influence of the Ampere force from the side of the constant magnetic field B ₀ .

Therefore, when comparing samples by the method of local spectroscopy, it is necessary to minimize their differences in the degree of destruction. In particular, it is necessary to set the same magnitude gradient values, and at the same time place the samples at the same distance from the center of the magnet. Otherwise, spatially selective pulses will have to set different fill frequencies and spectral widths, and then it will be impossible to ensure that RF excitation is completely identical for different samples.

The tomographic method is also known, which can potentially be used to study two samples located inside the scanning zone, i.e. to record a signal after RF excitation of two objects at the same time for different values of spatial coding of Larmor frequencies using gradient fields (Haacke EM, Brown RW, Thompson MR, Venkatesan R. Magnetic Resonance Imaging: Physical Principles and Sequence Design - John Wiley & Sons, 1999). The set of received signals is subjected to mathematical processing, which allows to obtain an MRI image. An analysis of the brightness distribution is performed for it. The pixel brightness (pixel = picture element) in this image is proportional to the intensity of the NMR signal from the corresponding voxel (voxel = volume element). Then the brightness of the pixels related to the images from samples A and B can be put in correspondence with the values of the NMR signals from them, obtained by the methods of conventional NMR spectroscopy at the very beginning of their registration, i.e. immediately after exposure to the RF pulse.

Unlike the method of local spectroscopy, this method does not need to apply the spin echo technique — you can use the gradient echo technique with a small FA (Frahm J., Haase A., Matthaei D. "Rapid NMR imaging of dynamic processes using the FLASH technique". Magnetic Resonance in Medicine. 1986. 3 (2): 321-327), respectively, can reduce the RF power requirements.

However, compared to conventional spectroscopic methods, MRI requires additional instrumentation. This applies to gradient coils and current sources for them, as well as equipment for the formation of complex-shaped pulses in order to obtain frequency-selective pulses. In addition, the method is difficult to implement in the event that the NMR spectrum contains not one but several lines. Then, in order to obtain information about each of them, it is necessary to carry out as many MRI scans, in which it is necessary to provide frequency-selective excitation of each line. The end result of using tomographic methods is a set of images, each of which is obtained by frequency-selective excitation of a separate line of the spectrum, where the brightness of each image is matched by the intensity of the NMR line obtained by conventional spectroscopic methods. Selective excitation can be difficult to implement in the case of overlapping adjacent spectral lines. Finally, information is needed on the position of the lines themselves in the spectrum.

In addition, the measurement requires more time than the usual spectroscopic method N times, where N is the size of the image matrix in the phase encoding direction for linear scanning or the number of signal recording sessions for radial scanning. And this applies to every line of the spectrum. If it is necessary to accumulate signals in order to increase the signal-to-noise ratio, the total scan time increases accordingly. With this approach, the total time of an MRI scan may be unacceptably long. In this case, it is not possible to view intermediate results until a sufficiently large number of scanning cycles take place, after which it is possible to reconstruct the image with a sufficiently high spatial resolution. In addition, the use of gradient pulses can cause a destructive effect on the studied samples, as in the method of local spectroscopy. Although, in contrast to the method of local spectroscopy, the destructive effect is the same for different samples, regardless of their location in the scanning zone, but this is not always acceptable.

In the prior art it is also known to use multi-coil systems (Array Coil) in MRI, with the help of which the scanning process of extended objects is accelerated due to simultaneous recording of signals from coils separated in space near the object under study (Parallel Imaging) - (Hutchinson M., Raff U. Fast MRI data acquisition using multiple detectors (Magn. Reson. Med. 1988, 6, p. 87-91). Each coil is connected to its receiver, and when reconstructing an image, the contribution of each of them is taken into account using a complex computational procedure (Carlson JW. RF coils. // J Magn Reson 1987: 74: 376-380 ).

For a multicolored system, each of the objects A and B can be placed in a separate section, and the registration of the NMR signals from them with can be performed simultaneously, but using two receivers. These signals can be assigned to the values of the NMR signals from them, obtained by the methods of conventional NMR spectroscopy at the very beginning of their registration.

The technical capabilities of tomographs equipped with the Array Coil system, in principle, allow not only tomographic scanning, but also the recording of ordinary spectra, and, simultaneously, for two samples. These tomographs could potentially be used for the simultaneous examination of two samples, provided that the samples are placed in different sections of the multicarrier system. Moreover, since the sections of the Array Coil coils are made so that their sensitivity zones partially overlap, in order to use them to obtain spectra from each of the samples, it is necessary to choose a system with at least 3 sections and place the studied samples in non-overlapping sections.

The disadvantage of this configuration is that all the coils and receivers to which they are connected, operate independently of each other. Usually the coils are part of the oscillating circuit with high Q. Violation of the resonance conditions for one of the circuits due to inaccurate tuning of the contour to resonance, as well as not identical characteristics for the receivers, significantly affects the measurement result, since identical conditions for recording signals for different samples are not provided. Non-identical measurement conditions for samples can be caused not only by instrumental imperfections or subjective factors of inaccurate contour tuning, but also by external factors that occur some time after the start of measurements, for example, temperature drift, which creates additional difficulties for fixing settings.

It should be noted that the application of the methods described above is not known for solving the problem of identifying an object by directly comparing the parameters of the test and reference samples using NMR signal registration using a two-coil circuit, which provides for the placement of test samples (identifiable and reference) in separate receiving coils, each of which has the same frequency defining elements for tuning the circuit containing this coil into resonance and connected to the same receiver with pos eduyuschim spectra comparison.

DISCLOSURE OF INVENTION

The technical problem solved by the claimed invention is the possibility of conducting non-invasive identification of objects by their NMR spectra using separate receiving coils and a common receiver with the possible use of common frequency defining elements.

The technical result of the proposed method of non-invasive identification of objects according to their NMR spectra and device for its implementation is to improve the accuracy of identification of the investigated object in comparison with the reference sample.

The invention extends the functionality of spectroscopic equipment, providing reliable registration of very small differences in NMR spectral characteristics in the presence of factors that significantly affect the measurement accuracy using traditional methods - instrumental imperfections, subjective factors when setting up equipment, equipment instabilities from temperature drift and other reasons.

To achieve the technical result, a device is proposed for detecting by NMR of samples differing in their characteristics from the reference one, which contains two series-connected inductors (hereinafter referred to as coils) made of the same material and the same parameters: coil diameter, number of turns, winding pitch, length and the wire thickness, while the coils are made with the possibility of placing in one of them a reference sample, in the second - the test sample, where one of the coils is connected at one end to Nome input ends of the signal receiver, the other - to the one end of the second coil, the remaining end of the second coil is connected to another input end of the receiver signal,

while the coils are made with the possibility of switching modes, in which the emf. inductions induced at the ends of the coils are summed or subtracted, respectively, the total or difference values enter the receiver input, and the device is arranged to be placed in a uniform magnetic field with a perpendicular arrangement of the coil axes relative to the direction of the field.

The device is made with the possibility of forming a resonant oscillatory circuit, the inductance of which is formed by series-connected coils, ensuring the matching of the resistance of the resonant oscillating circuit with the input resistance of the receiver. To form a resonant oscillating circuit, it contains the first capacitor connected in parallel with the inductance formed by the coils connected in series, and to ensure matching the input impedance of the receiver with the resistance of the oscillating circuit, it contains the second capacitor connecting the oscillating circuit with the receiver input.

To switch modes of operation, the device is equipped with a switch that can be made programmatically controlled.

With the help of this device is implemented a method of identifying samples that differ in their characteristics (physico-chemical) from the reference.

The method includes placing the reference and test samples in the specified device, followed by placing the coils with samples in a uniform magnetic field ensuring the coils axes are perpendicular to this field, and then radiofrequency excitation of the resonating spins is performed simultaneously in both samples, followed by recording NMR signals from both samples simultaneously on the specified excitation, first in the mode in which the emf inductions induced at the ends of the coils are summed, and then in the mode in which the emf inductions induced at the ends of the coils are subtracted, and a sample that differs in its characteristics (physico-chemical) from the reference one is detected by comparing the NMR spectra from the received signals.

The identification of a sample that differs in its characteristics (physico-chemical) from the reference sample is carried out according to the results of mathematical processing of the obtained spectra of S ⁺ and S ^- , where

S ⁺ - NMR spectrum from the signal recorded from the reference and test samples obtained from the mode of operation at which the emf. inductions induced at the ends of the coils are summed;

S ^- is the NMR spectrum of the signal recorded from the reference and test samples obtained from the mode of operation at which the emf inductions induced at the ends of the coils are deducted,

the spectra are calculated as S _A = (S ⁺ + S ^- ) / 2 and S _B = (S ⁺ -S ^- ) / 2, comparing which reveal differences in their characteristics between the reference and the studied samples.

To increase the signal-to-noise ratio, the registration of signals necessary to obtain the S ⁺ and S ^- spectra ^is repeated many times with the possibility of accumulating data, after which the values of S _A and S _{B are} calculated.

The claimed invention has several advantages.

The inventive method does not use inhomogeneous magnetic fields, therefore:

1) simpler technical equipment is used - no special coils and a current amplifier are required to create gradient fields, there is no need to use complex-shaped pulses with a narrow spectrum, reduced power requirements of the transmitter, since it is possible to work even with small FA;

2) it takes less time to measure; you can view intermediate results and reduce the measurement time by suspending the accumulation of signals if the achieved signal-to-noise ratio corresponds to a predetermined value;

3) there is no destructive factor associated with the effect of mechanical vibrations on the test samples, the requirements for their placement inside the magnet are simplified.

Compared to methods where non-uniform magnetic fields are not used, the claimed invention also has several advantages:

4) increases the accuracy of measurements, since simultaneous registration of a difference signal from two samples is used, which reduces the influence of instrumental instabilities on the measurement process,

5) the use of common for compared samples of the receiver and contour elements reduces the requirements for the accuracy of their settings and the identity of the manufacture of these hardware components.

Brief Description of the Drawings

The device is illustrated with graphic materials, where in FIG. 1-4 presents the variants of the electrical circuit, implemented by the claimed device, namely:

FIG. 1 - schemes of non-resonant and resonant circuits, where the voltage source is the emf of induction induced into the inductance by precessing nuclear spins;

FIG. 2 - schemes of resonant circuits, which include two inductors in series, each of which is a source of emf. induction induced in a coil by precessing nuclear spins. Two options for connecting the ends of the coils;

FIG. 3 - schemes of non-resonant circuits, which include two inductors in series, each of which is a source of emf. induction induced in a coil by precessing nuclear spins. Two options for connecting the ends of the coils;

FIG. 4 - variants of the device circuit - the basic variant, the variant, when the coils of the device are used in the composition of the resonant circuit, and the variant with the use of the switch, which allows to switch from the basic variant to the resonant circuit and back without dismantling the frequency-setting elements.

FIG. 5 shows a photo of the installation, which, using a manufactured device, tested the proposed method for non-invasive identification of objects by their NMR spectra;

FIG. 6 shows the arrangement of the claimed device in the base case together with the objects under study in the magnet gap.

Figure 7 presents a photo of a practically implemented device in the basic version with maximum detail.

On Fig presents the total and differential (relative to the emf induction induced in the coils) proton spectra from the studied objects - glasses of water, obtained from a practically implemented device in the base case.

FIG. 9 shows a photo of a practically realized device in a variant using a resonant circuit and a software controlled switch for recording NMR signals from an oxygen isotope ¹⁷ O.

FIG. 10 shows the total and differential (relative to the emf induction induced in the coils) ¹⁷ O NMR spectra in the objects under study - vodka bottles from different manufacturers, obtained in the device variant using a resonant circuit and a software controlled switch.

The implementation of the invention

Below is a detailed description of the proposed method and device for its implementation.

According to the claimed method, two stages of registration of NMR signals are produced successively - first register the total signal from two spaced apart samples placed in separate, but identically manufactured, coils - S ⁺ = S _A + S _B , and then differential - S ^- = S _A -S _B. From these data, calculate the spectra S _A = (S ⁺ + S ^- ) / 2 and S _B = (S ⁺ -S ^- ) / 2 and derivatives of them, in particular, S _A / S _B = (S ⁺ + S ^- ) / (S ⁺ -S ^- ) and (S _A -S _B ) / (S _A -S _B ) = S ^- / S ⁺ .

In the process of multiple accumulation of signals, it is possible to calculate the spectra from the results of partially conducted accumulation, and if a predetermined signal-to-noise ratio is reached, then the process can be suspended.

The useful effect is explained by the fact that with this method of comparing spectra from different samples, the instability of the instrumental parameters affecting the comparison result is much less than with the traditional method of recording NMR spectra from the same samples.

Below is a more detailed explanation of the achievement of this advantage.

Usually, NMR signals are recorded after pulsing a spin system using an RF field oriented perpendicular to a polarizing field B ₀ . When the pulse frequency is equal to the Larmor frequency ω = γB ₀ , where γ is the gyromagnetic ratio of the nucleus, the total vector of macroscopic nuclear magnetization deviates from the equilibrium state (parallel to the field B ₀ ), and after the end of the RF pulse, precession movement of the magnetization around this field is observed Larmor frequency.

The receiving coil is placed perpendicular to this direction, so that at its ends there is an electrical voltage - emf. induction. According to the Faraday law, it is equal to E = -dФ / dt, where Ф is the magnetic flux crossing the turns of the coil. In this case, this flow is Φ = MSNf, where M is the component of the macroscopic magnetic moment vector parallel to the axis of the coil, S is the area of the coil turn, N is the number of its turns, f is the coil filling factor (determined by the ratio of the areas of the coil turn and sample distribution within this area).

Since M varies according to the harmonic law - with the Larmor frequency, then the emf. varies with the same frequency. Therefore, the amplitude of the emf. is proportional to ωF = ωMSNf. The value of M is determined by the equilibrium value of the nuclear magnetization, the concentration of the resonating nuclei in the sample and the angle of deviation of the total magnetization vector from the B ₀ direction, achieved after the stimulating RF field is applied to the sample.

Induced in the coil emf E causes the electric current I in the circuit formed by the coil and its load - the resistance of the external circuit. In the simplest case, this resistance is the input impedance of the receiver R - the left fragment of FIG. one.

Then at the receiver input receive voltage S, proportional to the current, and, consequently, emf. induction E. Its amplitude value | S | is determined by the formula:

| S | = | E | × R / | Z |, where | Z | = ((R + r) ² + (ωL) ² ) ^1/2 , L is the inductance, r is the resistance of the coil.

For an infinitely large value of R, this value coincides with | E |. Otherwise, it is smaller.

To increase the signal at the receiver input, resonance methods are used - the coil is included in the oscillating circuit. To do this, parallel to the inductance L connect the capacitance C. If the condition LC = ω ^-1/2 is satisfied, then the voltage on the capacitor is S = EQ, where Q = ωL / r is the quality factor of the circuit. The value of Q can reach several hundred. Thus, when using a coil with a sample in the composition of the resonant circuit, emf induction is enhanced hundreds of times. If the resonance condition is not fulfilled, then the voltage dependence on the capacitor C on the frequency is more complicated:

| S _in | = | E | × (ωC) ^-1 / (r ² + (ωL-1 / ωC) ² ) ^1/2

In reality, the voltage at the receiver input is somewhat lower due to the need to match its input resistance (usually 50 ohms) with the resonant resistance of the circuit. For this purpose, an additional matching capacitor C _m is added to the circuit - the right-hand fragment of FIG. one.

Then the dependence of the signal at the receiver input from the emf. additionally complicated and depends not only on L and C, but also on C _m , as well as R. It is important that the signal is directly proportional to E:

| S _in | = | E | × K (L, C, C _m , R).

Further, icons indicating the magnitude module will be omitted.

If the receiver amplifies this signal P times, then the receiver’s output is finally

we have:

S _out = E × K (L, C, C _m , R) × P

To register the NMR signals using two unrelated identical coils L _A and L _B , each of which has a separate sample, each coil L _A and L _{B is} included in the resonant circuit by connecting capacitors C _A and C _B parallel to them which, in turn, through capacitors C _mA and C _{mB are} connected to their receivers with input resistances R _A and R _B , amplifying the signal by P _A and P _B times, respectively. Then at the output of the receivers receive:

S _A ⁼ E _A × K (L _A , C _A , C _mB , R _A ) × P _A and S _B = E _B × K (L _B , C _B , C _mB , R _B ) × P _B.

If the samples are identical (with the same RF exposure, they give the same response, and therefore the same emf of the induction), the parameters of the circuits are the same (the same inductance and capacitance values), the same technical characteristics of the receivers (the same input resistances and amplifications), There will be no difference in the signals S _A and S _B.

If the samples are not identical, but the parameters of the circuits and receivers are identical, then differences of the samples can be identified by comparing the values of S _A and S _B.

If there is non-identity for at least one pair of parameters, and all of them - 5 _{(L A, L B),} (C A, C B), (C mA, C mB), (R A, R B), (P A , P _B ) the result of the sample comparison proves to be unreliable.

The measurements will be unreliable even if at the beginning of the measurements the settings of the circuits and the parameters of the receivers were identical, but later, due to temperature drift and other instrumental factors, the parameters for at least one of them changed.

When using the claimed device, the requirements for identity relate only to one pair of parameters - L _A and L _B. This is explained as follows.

The parameters of the oscillating circuit in the inventive device are determined by two inductances L = L _A + L _B and the total capacitance C. Its resonant frequency is determined by the formula: LC = ω ^-1/2 . A current generated in the circuit of this circuit is formed, depending on the switch position, by the sum or difference of the emf. induction from each of the samples. And the voltage on the common frequency setting capacitor C through the common matching capacitor C _{m is} fed to the input of the common receiver. FIG. 2 shows the equivalent electrical circuits, implemented with two ways to connect the ends of the coils, where as the source of emf. need to understand emf induction from precessing magnetic moments. Emf induction varies according to the harmonic law, so the polarity of the voltage is indicated for the same point in time. With the same method of RF excitation of samples placed in separate identical coils, the phase of the emf, and, consequently, the polarity at the ends of the equivalent voltage sources at any time, coincide and do not depend on the method of connecting these coils.

Therefore, with two ways of connecting the ends of the coils at the output of the receiver, it turns out:

S ⁺ = (E _A + E _B ) × K (L, C, C _m ) × P and S ^- = (E _A -E _B ) × K (L, C, C _m ) × P.

Then the values used to measure sample differences, in particular, S _A / S _B are independent of L, C, C _m and P. For a value of S _A / S _B , which is equal to (S ⁺ + S ^- ) / (S ⁺ -S ^- ), we have: S _A / S _B = (E _A + E _B ) / (E _A- E _B ). That is, these values are determined only by the differences in the emf generated by the samples, and does not directly depend on the instrumental factors. This is due to the fact that the instrumental factors affecting the recorded NMR signal (loop tuning, matching with the receiver, receiver gain) equally affect the signals from both samples. Therefore, the ratio of the total signal to the difference signal at the output of the receiver remains unchanged.

In the event that there is no need for large amplification of the NMR signal or resonant elements embedded in the receiver, then capacitors are removed from the device, and it works as a non-resonant circuit. Electrical circuits of the device with two methods of connecting the coils for this case are shown in FIG. 3

The variability of each of the values of E _A and E _B separately depends little on the variability of instrumental factors. This follows from the formula for the emf. E ~ ωMSNf. Even if we assume that for some reason (the effect of temperature) the coil area S can change, then the filling factor f will change in inverse proportion, and at constant sample sizes, the value of E will remain unchanged.

However, the parameters of the coils L _A and L _B must be identical. This concerns, first of all, the number of turns, since this parameter directly affects the emf value. The identity of the remaining parameters is necessary so that the values of the inductances L _A and L _B are equal at any moment in time. Their change is allowed, for example, under the action of temperature, but these changes must be the same for L _A and L _B. Otherwise, the conditions for the registration of NMR signals for the studied samples located in these coils will be nonidentical.

This is explained as follows. In the above calculations did not take into account the fact that the current induced from the emf. induction, creates in the coil a secondary magnetic field acting on the backs in such a way as to force their return to the equilibrium state - the effect of so-called. radiation damping (Bloembergen N., Pound RV Radiation damping in magnetic resonance experiments // Phys. Rev. 1954; 95: 8-12). This effect leads to a weakening of the recorded signal and the broadening of spectral lines. The value of the secondary field H ₁ inside the coil with inductance L and the area of the coil S can be estimated by the formula LI = H ₁ S, where I is the current circulating in the circuit, due to the emf. induction. Thus, the secondary field is directly dependent on the magnitude of the inductance and the area of the coil coil. Although the effect of radiation attenuation is small, however, for precision measurements it is necessary to ensure the conditions under which it is the same for the compared samples. Since the same current I flows through the coils, for the equality of the field values H _{1 it is} enough that the parameters of the coils are the same.

It should be noted that the coils that are part of the claimed device can be used not only for recording NMR signals, but also for RF excitation of a spin system. This is possible because the processes of spin initiation and registration of the response from them are separated in time. This method of using the same coil, both for receiving a signal and generating a RF field that excites spins (Single Coil circuit), is common in NMR. Therefore, in typical NMR equipment, a receive / transmit switch is installed at the receiver input, with which the coil ends fed to the receiver can be switched to the transmitter output. In this case, there is no need for a separate transmitting coil, which should generate a uniform RF field B ₁ in the zone where the test samples are located (Cross Coil scheme). If the transmitting coil is not large enough in the Cross Coil scheme (commensurate or smaller than the objects under study), then it becomes difficult to achieve the same RF field for two samples spaced apart in space, and the requirements for the location of the objects are tightened - they should be located symmetrically relative to its center.

A variant of the Single Coil scheme, in which the coils of the claimed device are used to excite NMR signals from different samples, is interesting in that the equality of the RF fields for each sample is ensured automatically. This is due to the fact that their values are determined by the equality LI = B ₁ S, where I is the current generated by the transmitter. But since the parameters of the coils - the inductance L and the area of the coil S are the same, and the coils themselves are connected in series, the current I is the same for both coils, and consequently, the values of B _{1 are the} same.

The disadvantage of the Single Coil scheme is the different size of the RF field inside the coil with the sample - it decreases with distance from its axis. Because of this, the total NMR signal from the sample is less than in the case of Cross Coil.

The specificity of using device coils in the Single Coil scheme is that RF excitation should always occur with the same variant of connection of these coils. This is due to the fact that the phase of the emf. induction is determined by the phase of the field B ₁ , and for the two variants of the connections of the coils of the phase the fields generated by the coils differ by 180 degrees. If the coil connections are set the same both when excited and when receiving a signal, then the S ⁺ values will be equal to S ^- , and comparison of the spectral parameters by the claimed method will be impossible. Therefore, if the device coils are supposed to be used not only to receive NMR signals, but also to generate a field, it is necessary not only to activate a receive / transmit switch in the receiver, but also to provide for the installation of the same type of coil connection in transmission mode.

Thus, the inventive method reveals much more reliably differences in samples of magnetic resonance characteristics, since measurements are made simultaneously, a signal is recorded that directly reflects these differences, and since the measurements use the same receiver and the same frequency-defining elements , the variations of their parameters do not directly affect the result of measurements aimed at identifying differences in samples. It is important that the claimed method requires the use of only one receiver instead of two.

An indirect factor affecting the measurement error is the accuracy of tuning the circuit for resonance and matching it with the receiver, because if there is inaccurate tuning and matching, the signal at the receiver input decreases at a constant noise level. Since the accuracy of these settings is not always monitored due to hardware imperfections and the impact of destabilizing factors (for example, temperature drift), it is useful to monitor the signal-to-noise ratio and, if necessary, increase the number of signal accumulations.

The method can be implemented using a device whose circuit is shown in FIG. 4. The left part shows the basic part of the device - two identical inductors (L _A and L _B ), inside which load the compared samples A and B, and switch S ₁ , switching the ends of these coils, and two wires connecting the device to the receiver.

In the device, it is possible to connect between the base part and the receiver of frequency setting elements, providing the resonant properties of an oscillating circuit, the inductance of which is composed of series-connected inductances L _A and L _B. In addition, it is possible to connect elements to such a contour to match the resistance of the circuit with the input impedance of the receiver (usually 50 Ohms). In the central portion of FIG. 4 shows one of the options for connecting these elements to the base part of the device - the frequency setting capacitor C and the matching capacitor C _m .

In the right fragment of FIG. 4 shows a variant of the device circuit that allows using the switch S ₂ to disconnect the capacitors without dismantling them, as a result of which the device can operate as if only its basic part is involved. Such a scheme can be required if there is no need for a significant amplification of the NMR signal, or if the frequency setting capacitor C and the matching capacitor C _{m are} already installed in the receiver.

The device is installed so that the test samples are in the zone of a uniform magnetic field (working zone) of the tomograph or spectrometer with a working gap sufficient to accommodate them. Usually the working area is located near the geometric center of the magnet. The axes of the coils are oriented perpendicular to the polarizing magnetic field.

In the position of the switch I (mode I), the ends of coils A ₂ and B ₂ are interconnected, and the remaining ends A ₁ and B ₁ are used to connect the device to the receiver. In this case, the input of the receiver receives a voltage proportional to the total NMR signal from two samples (the sum of the emf induced by two coils).

In the position of switch II (mode II), the ends of coils A ₁ and B ₂ are connected, and the receiver is connected to ends A ₂ and B ₁ . In this case, the input of the receiver receives a voltage proportional to the difference NMR signal from the two samples (the difference in emf of the induction from the two coils).

The NMR signals from the samples occur after pulsed exposure to RF pulses. Those. are the response of the spin system to this effect.

By registering the response of the spin system at different positions of the connections of the ends of the coils (or switch positions), you can register either the sum of the NMR signals from objects located in different coils, or their difference. Signals vary in time in a complex way, so it is convenient to present these total and difference signals as spectra.

If to obtain spectra it is necessary to increase the signal-to-noise ratio, then the accumulation of S ⁺ and S ^- signals ^is performed with their alternation. In this case, to switch the ends of the coils, it is advisable to use a program-controlled switch, for which various structural solutions are known - electronic circuits, for example, using switching pin diodes, etc.

Examples of the implementation of the invention

To implement the proposed method, two versions of the device were manufactured, which were tested on a 0.5 T MP Bruker Tomikon S50 tomograph (hereinafter referred to as MRI) - FIG. 5. Larmor's frequency of protons - 21.08 MHz. MRI contains a transmitting coil 1 with a diameter of 60 cm, rigidly fixed in the cavity of the superconducting magnet 2. Coil 1 is needed to generate RF field B ₁ due to the current supplied to it from an external generator. This field is necessary for the excitation of nuclear spins, and it is created by currents flowing through the annular conductors placed on the sides of the cylindrical surface of coil 1. Therefore, field B _{1 is} directed in the horizontal direction perpendicular to the polarizing field B _o . FIG. 5, the direction of the field B _{0 is} indicated by a thickened arrow.

The superconducting magnet 2 is made in the form of a current solenoid immersed in a 1200-liter tank filled with liquid helium. The induction of the static field of the magnet 2 in its working part within a sphere with a diameter of 40 cm is 0.5 T. The room in which magnet 2 is located is shielded using copper plates to minimize the effect of external RF radiation on MRI. Electronic units are installed in a separate room - a processor that generates tomograph control signals, a transmitter and a receiver that receives the NMR signal from the preamplifier 6, with its subsequent digitization and recording in electronic form. The tomograph control signals — RF and gradient pulses go to power amplifiers and then to MRI from a separate room where the control unit is located. It is based on an operator-controlled computer. Using this computer, instructions are sent to the processor and data about the digitized signal is received. On the same computer, which is controlled by the operator, data processing takes place - reconstruction of MRI images and their post-processing processing - scaling, contrast control, volume processing, archiving by writing to a disk or other media, issuing images on film or as a printout.

Both versions of the device have the ability to be placed inside the cavity of magnet 2. Its location in the direction of the polarizing field B _{0 is} controlled by the mobile platform 4. Using the RF cable 5, the signal from the device is fed to the pre-amplifier 6.

The upper surface of the mobile platform is made with a downward deflection, which prevents a stable arrangement of the device. Therefore, on top of the platform, there is a housing from a proprietary flat coil 3, the upper part of which has a flat section. The device was located on this site. Since the device performs the functions of the receiving coil, it is connected by cable 5 to the input of the preamplifier 6.

The functions of the receiving coil were performed by the inventive device, by means of which the comparison of samples was carried out, which were visually identical vessels filled with water.

The above MRI equipment is typical and does not require any modification to implement the proposed method. FIG. 6 shows the device with samples in a larger view, showing the location of the device in the simplest configuration (using only the base part) together with the objects under study in the gap of the magnet 2.

FIG. 7 shows a more detailed photo of a practically implemented device for recording proton signals at a frequency of 21.08 MHz. Inside the single-turn coils 7 and 8 (L _A and L _B , according to Fig. 4), the samples under study (A and _B , according to Fig. 4) are placed with glasses of water. The wire thickness is 2 mm, the diameter of the coil is 6 cm. The ends of the coils go to switch 9. One of the switched ends (A1 or A2, as shown in Fig. 4) goes to the switch on two-wire cable 5, to the other end of the cable goes one of the coil ends , which, according to the scheme (Fig. 4), where it is designated as B ₁ does not commute. Since in this case the signal / noise is greater than 100, there was no need for a prolonged accumulation of signals. Therefore, only the basic part of the device was used (the left fragment of Fig. 4), and a mechanical device — a two-section toggle switch — was used as a switch.

The axis of the coils (their diameter is 6 cm) is oriented perpendicular to the direction of the field B ₀ . To excite the NMR signal, an RF pulse with a duration of 100 μs was applied to the transmitting coil 1, after which a proton signal from water was recorded. Fourier signal processing gives a spectrum in the form of a single peak - FIG. eight.

FIG. 8 shows the results of studies of the samples depicted in FIG. 7

Here are the spectra of S ⁺ (upper spectrum) and S ^- (lower spectrum). They are obtained at different positions of the switch. The zero lines of the spectra are shifted in the vertical direction for the convenience of visual perception. The difference spectrum S ^- has the form of a dispersion (S-shaped) curve due to the presence of a small inhomogeneity of the polarizing field along the direction connecting the centers of the coils. Integrating the spectra allows to level this effect and estimate the values of S _A and S _B by the values of the integrals measured for S ⁺ and S ^- . In this case, these integrals are 1.0 and +0.02138. Then S _A / S _B = (S ⁺ + S ^- ) / (S ⁺ -S ^- ) == 1.02138 / 0.97862 = 1.04.

This means that the signal from sample A is 4% larger than that from sample B. The reason for this in this case is the difference in the volume of the samples, which is unnoticeable, but revealed by weighing objects A and B.

FIG. 9 shows another example implementation of the device. It is adapted for recording NMR signals from weakly sensitive nuclei - ¹⁷ O. Its Larmor frequency is 2.87 MHz, which is much less than that of a proton. Therefore, the number of turns for the coils is increased to 25, and the turns themselves (wire thickness 2 mm) are evenly distributed along the height of the cardboard cylinder (diameter 7.8 cm, height 11.5 cm). Elements C and C _m are added to the base part of the device, respectively, 10 and 11, necessary for the device to work as a resonant circuit. These are ceramic rotary capacitors of variable capacitance - from 8 to 30 pF. In addition, a software-controlled switch is introduced based on the electromagnetic relay 12. It is used to connect the ends of the coils in the same way as for the manual switch, the circuit of which is shown in FIG. 4. Mode I or II is selected by applying a 12 V voltage to the coil of the relay - in the absence of voltage, the device operates in I mode, and when 12 V is applied it is in mode II. The control voltage is supplied to the relay by means of a cable 13 connected to a device capable of issuing 12 V pulses upon commands from a pulse programmer.

In this tomograph, as in most typical models, there are electronic resources and software tools to perform mode switching in the process of recording NMR signals.

To ensure that the constant magnetic field of the tomograph does not interfere with the operation of the relay, it is provided to be placed outside the magnet gap, for which purpose the relay is connected to the coils of the device using additional conductors connected to the contact pad 15. For the convenience of operating the device, cables connected to it (signal and control relays) are made shielded, short - 50 cm with BNC type connectors - 16.17 for connecting them with much longer cables - 4 m connecting the device with tomograph nodes - preamplifier and the output of the pulse programmer.

For comparison of samples - two bottles of vodka with a capacity of 0.25 l with a strength of 40% produced multiple accumulation of 17O NMR signals with alternation of modes I and II. A signal was first recorded in mode I, then in mode II. Then - again in mode I, and then in mode II, etc. The results of recording the NMR signals obtained in mode I were summed, and the sum was subjected to Fourier processing, which made it possible to obtain the S + spectrum. Similarly, the registration results obtained in mode II were summed up, followed by obtaining the S- spectrum.

FIG. 10 shows the spectra of S + (upper spectrum) and S- (lower spectrum). The zero lines of the spectra are shifted in the vertical direction for the convenience of visual perception. The difference spectrum has a clearly pronounced peak, which indicates that the signal from sample A is greater than from B. The reason for this in this case is the difference in transverse relaxation times T2, due to which the induction signal decays at different speeds, which is reflected in spectral line intensity. T2 differences themselves are usually associated with the presence of additives or impurities. It is convenient to detect the difference in T2 from the usual 17O spectra, rather than proton ones, since for the 17O nucleus the lines are wide> 100 Hz, and to measure T2 it is enough to compare the width of the spectral lines. The lines of the proton spectra are narrow — their width is less than 1 Hz, while the inhomogeneous broadening due to the insufficient uniformity of the B0 field in the tomograph is several times larger. Therefore, it is impossible to identify differences in T2 in the proton spectra, and more complex measurements are required — using the multiple spin echo technique.

To clarify the reasons for the lower mobility of the molecules, and, consequently, the shortening of the time T2, MRI studies of the samples were carried out on a high-field (600 MHz) NMR spectrometer Bruker Avance 600. For this, part of the contents of the bottles was placed in ampoules 5 mm in diameter. Analysis of the proton NMR spectra obtained as a result of a long-term (30 min.) Signal accumulation revealed that the amount of impurities in the sample under study was 15% higher than in the reference sample. This may well be the reason for reducing the relaxation time T2, and, consequently, reducing the height of the peaks in the 17O spectra. Thus, the differences in the spectral parameters of the samples revealed by the non-invasive express-method (5 minutes) were confirmed by invasive detailed (30 minutes) research.

Claims (11)

1. A device for detecting by NMR of samples that differ in their characteristics from the reference one, including two series-connected identical inductors, connected to a signal receiver and configured to accommodate a reference sample in one of them, in the second test sample, one of the coils one end connected to one of the input ends of the receiver signal, the other to one end of the second coil, the remaining end of the second coil connected to the other input end of the receiver signal la, coils are made with the possibility of switching modes of operation in which the induced emf induced at the ends of the coils are added or subtracted, and the device is arranged to be placed in a uniform magnetic field with the perpendicular arrangement of the coil axes relative to the direction of the field. 2. The device according to claim 1, characterized in that it is made with the possibility of forming a resonant oscillating circuit, the inductance of which is formed by series-connected coils, with ensuring matching of the resistance of the resonant oscillating circuit with the input impedance of the receiver. 3. The device according to claim 2, characterized in that for the formation of a resonant oscillating circuit it contains the first capacitor connected in parallel with the inductance formed by the series-connected coils, and to ensure matching the input impedance of the receiver with the resistance of the oscillating circuit it contains the second capacitor connecting the oscillating circuit with receiver input. 4. The device according to claim 1, characterized in that for switching modes of operation it is equipped with a switch. 5. The device according to claim 4, characterized in that the switch is made program-controlled. 6. A method of identifying samples that differ in their characteristics from the reference one, including placing the reference and test samples in the device according to claim 1, followed by placing the coils with the samples in a uniform magnetic field ensuring the perpendicular arrangement of the axes of the coils to this field, then using NMR perform radio frequency excitation of resonating spins simultaneously in both samples with the subsequent recording of NMR signals simultaneously from both samples to the specified excitation, first in the mode where the induction emf induced at the ends of the coils are summed, and then in the mode in which the induction emf induced at the ends of the coils are subtracted, and a sample that differs in its physico-chemical characteristics from the reference one is detected by comparing the NMR spectra from the received signals . 7. The method according to p. 6, characterized in that the identification of the sample, which differs in its physico-chemical characteristics from the reference sample, carried out according to the results of mathematical processing of the obtained spectra S ⁺ and S ^- , where S ⁺ is the NMR spectrum of the signal recorded from the reference and test samples obtained from the mode of operation at which the emf. inductions induced at the ends of the coils are summed; S ^- - NMR spectrum from the signal recorded from the reference and test samples obtained from the mode of operation in which the induction emf induced at the ends of the coils are subtracted, the spectra are calculated as S _A = (S ⁺ + S ^- ) / 2 and S _B = (S ⁺ -S ^- ) / 2 and the differences in their characteristics between the reference and test samples are determined by comparing the obtained spectra. 8. The method according to claim 7, characterized in that the registration of signals to obtain the spectra S ⁺ and S ^{- is} repeated many times with the possibility of accumulating data to increase the signal-to-noise ratio, and then calculate the values of S _A and S _B.

Images